If the character and causation of fatigue following highly skilled work are to be understood, the first need is for the discovery of more relevant and experimentally controlled facts. Unfortunately, almost all the investigators who have attempted to study fatigue of this type have adopted methods taken over with very slight change from those which have proved valuable in the study of simple muscular fatigue. They have chosen elementary operations usually considered to require some 'mental' effort--such as easy calculations, word or colour recognition and naming and the like--have repeated these operations over and over again for long periods, and have tried to express the resulting fatigue in terms of the diminution in quantity or quality of the work done. The skill fatigue of daily life is not set up under such conditions. Routine repetition of simple actions is not a characteristic of any highly skilled work, and least of all of work having a strong 'mental' component. The operations involved here are marked by complex, co-ordinated and accurately timed activities. The stimuli in response to which these activities are set up are neither simple nor do they usually fall into an order of fixed succession. They have the character of a field, or a pattern, which has become very highly organized, and may retain its identity in spite of a great diversity of internal arrangement. It is possible to develop fully controlled experimental situations in which these realistic considerations have full play. When this is done the picture of fatigue following highly skilled work which emerges has certain strongly marked characters. In such fatigue the 'standards' accepted and followed by the central nervous system unwittingly deteriorate. The operator tends to think that he is doing better work, because errors treated as significant all the time get wider and wider limits. Until a stage of great fatigue is reached, it is far more likely that the right actions will be performed at the wrong times than that the wrong actions will be performed. If accurate timing is insisted upon, gross mistakes of action may appear. The stimulus fields splits up. Its pattern character alters. It becomes a collection of unconnected signals for action, with some of these predominant over all the others. Particularly, stimuli which are in the margin of the pattern, not closely organized with the central field, are ignored, 'forgotten', and serious lapses of specific reactions occur. There is a marked change in the effect of certain 'distracting', or additional stimuli. Sensations of bodily origin, in particular, become more pressing and insistent and affect the performance in ways peculiar to the tired operator. Side by side with all these changes go constant subjective symptoms. Verbal reports about any circumstances connected with known failure of performance become increasingly inaccurate, and errors are regularly projected upon objective conditions, or attributed to the interference of other people. There is a tremendous growth of irritability. An attempt is made to discuss the light thrown by this picture upon the relation of high-level central nervous functions to simpler neuro-muscular mechanisms.